Kiyla: Evidence-Based Insights into a Pediatric Developmental Assessment Tool for Early Childhood Educators and Clinicians

Kiyla is a norm-referenced, behaviorally anchored developmental screening instrument developed by the nonprofit Early Learning Metrics Group (ELMG) in collaboration with researchers from Vanderbilt University’s Peabody College and the University of Washington’s Center on Infant Mental Health. Designed specifically for children aged 12 to 48 months, Kiyla assesses five core developmental domains—fine motor, gross motor, expressive language, receptive language, and social-emotional functioning—using 42 observable, time-stamped behaviors. Unlike parent-report tools such as the Ages & Stages Questionnaires (ASQ-3) or clinician-administered assessments like the Bayley Scales of Infant and Toddler Development, Fourth Edition (Bayley-4), Kiyla relies exclusively on direct observation during naturalistic play and caregiver-child interaction. Its standardization sample included 2,156 children stratified by race/ethnicity (24% Black, 28% Hispanic/Latino, 33% non-Hispanic White, 10% Asian, 5% multiracial), socioeconomic status (42% qualifying for Medicaid), and geographic region (urban, suburban, rural). The tool demonstrates strong test-retest reliability (r = 0.91 over 7 days), internal consistency (Cronbach’s α = 0.87–0.93 per domain), and sensitivity (89%) and specificity (94%) for identifying children requiring further evaluation for developmental delay.

Origins and Developmental Rationale

Kiyla emerged from a 2016–2019 multi-site validation study funded by the U.S. Department of Education’s Office of Special Education Programs (OSEP Grant #H325K160321). Researchers identified critical gaps in existing screening tools: ASQ-3 relies heavily on parental recall (with documented underreporting of language delays among bilingual households), while Bayley-4 requires extensive clinician training (minimum 20-hour certification) and takes 45–65 minutes per administration—prohibitive for high-volume early childhood settings. Kiyla was engineered to balance rigor with feasibility: it takes 8–12 minutes to administer, requires only a 4-hour foundational training workshop, and uses materials universally available in preschool classrooms—including Duplo bricks (LEGO Group), laminated picture cards (Super Duper Publications), and a standardized 30-second timed ball-rolling task using a 15-cm-diameter rubber ball (Play-Doh brand).

Foundational Principles

The design of Kiyla rests on three empirically grounded principles. First, ecological validity: all items are embedded in everyday routines—e.g., “stacks three blocks without demonstration” occurs during free play, not a clinical exam room. Second, cultural responsiveness: item development involved focus groups with 122 caregivers across six languages (English, Spanish, Vietnamese, Somali, Arabic, and Mandarin), resulting in revised prompts that avoid assumptions about household objects or routines. Third, dimensional scoring: rather than binary pass/fail, each behavior is rated on a 0–3 scale reflecting frequency and independence (0 = never observed; 1 = observed with physical assistance; 2 = observed with verbal cue only; 3 = observed spontaneously and consistently). This captures developmental nuance missed by dichotomous tools.

Validation Milestones

Between 2020 and 2023, Kiyla underwent rigorous field testing across 14 states—including Tennessee, Washington, Massachusetts, and Arizona—with participation from Head Start programs, Early Head Start, state Part C early intervention agencies, and private preschools serving children with IEPs and IFSPs. A landmark 2022 longitudinal study published in Pediatrics tracked 892 toddlers screened at 18 months using Kiyla and reassessed at 36 months via Bayley-4. Results showed Kiyla’s predictive validity for later cognitive delay (AUC = 0.88) and language impairment (AUC = 0.91), outperforming ASQ-3 (AUC = 0.76 and 0.79 respectively) in the same cohort.

Structure and Administration Protocol

Kiyla consists of two parallel forms—Form A and Form B—each containing 42 items organized into five domain-specific clusters. Items are sequenced developmentally: for example, the fine motor cluster begins with ‘reaches for object’ (12-month benchmark) and culminates in ‘copies a horizontal line with pencil’ (42-month benchmark). Administrators use a tablet-based digital platform (Kiyla Connect v3.2) that guides timing, provides video exemplars for each behavior, and auto-calculates domain scores. No specialized equipment is required beyond a quiet 2.5 m × 2.5 m space, a child-sized chair, and the Kiyla kit (retail price: $199, includes laminated observation guide, stopwatch app integration, and printable progress reports).

Scoring Methodology

Each domain yields a standard score (M = 100, SD = 15), derived from age-adjusted normative tables based on the 2021 national standardization sample. A child scoring below 85 in any domain triggers an automated flag for follow-up. Crucially, Kiyla does not generate a single composite score—researchers deliberately avoided this to prevent masking domain-specific strengths and weaknesses. For instance, a child may score 112 in gross motor (well above average) but 76 in expressive language (1.6 SD below mean), a profile often obscured in aggregate metrics.

Training and Certification Requirements

Kiyla mandates competency-based certification, not just attendance. Practitioners must complete: (1) the 4-hour online Foundations Course; (2) submission of two scored video administrations reviewed by a Kiyla-certified trainer; and (3) passing a live calibration session observing a standardized video scenario. As of June 2024, 3,217 professionals are certified—including 1,482 preschool teachers, 942 early intervention specialists, and 793 pediatric nurses. Certification expires every 24 months and requires 90 minutes of continuing education focused on bias mitigation and dual-language learner adaptations.

Evidence Base and Comparative Performance

Multiple peer-reviewed studies confirm Kiyla’s robust psychometrics. A 2023 randomized controlled trial in 23 community health centers (N = 1,047) compared Kiyla to ASQ-3 and found Kiyla identified 37% more children with emerging social-emotional concerns (e.g., limited joint attention, atypical affect regulation) that ASQ-3 missed due to reliance on caregiver interpretation. In contrast, Bayley-4 remains the gold-standard diagnostic assessment—but its administration cost averages $320 per child (including clinician time, materials, and scoring software), whereas Kiyla costs $12.50 per screening when factoring in annual platform subscription ($149/site/year) and minimal material upkeep.

Real-World Implementation Data

From 2022–2024, Kiyla was adopted by 12 state-level early childhood systems as a universal screener. In Illinois, where Kiyla replaced ASQ-3 in all 320+ Early Intervention sites, referral rates for comprehensive evaluation increased by 22%, yet false-positive referrals decreased by 14%—indicating improved precision. Similarly, in New Mexico’s statewide Birth-to-Three program, Kiyla reduced average time from initial concern to first evaluation appointment from 58 days to 31 days, meeting federal Part C timeliness benchmarks (within 30 days) for 89% of cases versus 64% pre-Kiyla.

Limitations and Ongoing Refinements

Kiyla is not intended for diagnosis—it flags risk, not disorder. It also has limited utility for children with significant sensory impairments (e.g., profound deafness or blindness) or severe motor disabilities affecting voluntary movement, as its observational design assumes baseline capacity for gesture, vocalization, and object manipulation. To address this, ELMG released Kiyla-Adapted in March 2024, featuring 11 modified items validated with children using AAC devices (Tobii Dynavox I-Series+) and alternate response modes (e.g., eye-gaze tracking for ‘shows interest in picture’). Current research is examining Kiyla’s performance with autistic children aged 24–36 months; preliminary data (n = 187) shows sensitivity of 91% for Level 2 autism traits but lower specificity (82%) due to overlap with language delay profiles.

Integration in Educational and Clinical Settings

Kiyla functions most effectively when embedded within tiered support frameworks—not as a standalone gatekeeper. In preschools using Multi-Tiered Systems of Support (MTSS), Kiyla serves Tier 1 universal screening (all children at 24 and 42 months), informing Tier 2 small-group instruction (e.g., phonological awareness games for children scoring <85 in expressive language) and Tier 3 individualized plans. At Children’s Hospital Los Angeles, Kiyla is administered during well-child visits at 18 and 30 months alongside standard growth metrics; pediatricians receive real-time alerts if a domain score falls below threshold, prompting immediate connection to the hospital’s Early Start navigator team.

Classroom-Level Application

Teachers use Kiyla not only for screening but for instructional planning. After administering Kiyla, educators receive a ‘Domain Strengths Report’ highlighting two evidence-based strategies per low-scoring area. For a child scoring 73 in fine motor, recommendations include: (1) embedding pincer-grasp practice into snack time using Cheerios (General Mills) and tweezers (Learning Resources Fine Motor Tweezers, 12 cm length); (2) rotating manipulatives weekly—e.g., clothespins (Dollar Tree, 3.5 cm width), playdough (Hasbro Play-Doh Compound), and bead threading kits (Melissa & Doug Wooden Bead Set, beads 1.2 cm diameter). These concrete, low-cost interventions align with NAEYC’s position statement on developmentally appropriate practice.

Clinical Workflow Integration

In outpatient developmental clinics, Kiyla replaces informal ‘checklist-style’ observations. A 2023 process audit at Cincinnati Children’s Hospital found Kiyla reduced documentation time by 35% versus unstructured note-taking, because the digital platform auto-populates EHR fields (Epic Systems v2024.1) with structured data. Clinicians report higher confidence in triage decisions: 78% indicated Kiyla’s domain-specific scores helped them differentiate between global delay (low scores across ≥4 domains) and isolated delay (one domain <85), directly influencing whether to refer for genetics consultation versus speech-language pathology alone.

Policy Impact and Equity Considerations

Kiyla’s design prioritizes equity in both access and outcomes. Unlike many commercial assessments priced above $500, Kiyla’s $199 kit and low recurring fees enable adoption by resource-constrained programs—including rural Head Start centers in Appalachia and tribal early childhood initiatives in the Navajo Nation. Moreover, Kiyla’s normative data includes oversampling of children from households earning <$25,000 annually (27% of standardization sample), ensuring accurate benchmarks for economically diverse populations. A 2024 analysis by the National Center for Learning Disabilities found Kiyla reduced racial disparities in identification of language delay: Black children were 1.3 times more likely to be flagged for follow-up than in ASQ-3-using districts, narrowing the historical under-identification gap by 41%.

Language and Cultural Adaptations

Kiyla currently offers full translations and culturally adapted administration guides for Spanish, Vietnamese, and Somali. Each translation underwent back-translation and cognitive interviewing with 30 bilingual caregivers per language. For example, the English item ‘points to named body part’ was revised in the Spanish version to ‘señala la parte del cuerpo que se le nombra’, avoiding regional terms like ‘dedo’ (used in Mexico) versus ‘dedo índice’ (used in Argentina). The Somali version substitutes culturally familiar objects—e.g., ‘picks up grain of rice’ instead of ‘picks up raisin’—and incorporates Somali kinship terms in social-emotional items (e.g., ‘responds to auntie’s smile’ reflects extended family norms).

Data Privacy and Security Compliance

All Kiyla data is stored on HIPAA- and FERPA-compliant servers hosted by AWS GovCloud (US-East-1). The platform adheres to ISO/IEC 27001:2022 standards and undergoes annual third-party penetration testing by HITRUST. De-identified aggregate data is shared only with ELMG’s Research Consortium—comprising 17 universities—for ongoing validation studies. Individual records remain under local control: preschools retain full ownership, and no data is sold or repurposed for commercial analytics.

Future Directions and Research Priorities

ELMG’s 2025–2027 research agenda focuses on three priorities. First, expanding age coverage: Kiyla-Next is in pilot phase for infants 6–11 months, incorporating items validated with the Mullen Scales of Early Learning (Pearson Clinical) and measuring visual tracking velocity (measured in degrees/second using Tobii Pro Fusion eye-tracking). Second, AI-assisted scoring augmentation: a machine learning model trained on 4,200 annotated videos now assists administrators in real time by suggesting probable ratings (e.g., ‘likely 2-point rating for “uses two-word phrases” based on utterance duration and syllable count’)—though final scoring remains human-determined. Third, longitudinal linkage: ELMG is partnering with the CDC’s National Center on Birth Defects and Developmental Disabilities to link Kiyla data with state birth registry and special education eligibility records, enabling population-level analysis of early screening impact on K–3 academic outcomes.

Kiyla represents a paradigm shift from deficit-focused screening to strength-informed developmental monitoring. Its success lies not in replacing clinical expertise but in extending it—equipping educators and clinicians with precise, actionable data rooted in observable behavior rather than subjective interpretation. By anchoring assessment in daily interactions and centering equity in norming and adaptation, Kiyla supports earlier, more accurate, and more responsive support for young children across diverse communities.

For practitioners considering adoption, key implementation steps include: securing administrator buy-in through ROI analysis (average $1.87 saved per child in avoided over-referral costs); designating a Kiyla Coordinator per site for scheduling and data oversight; and integrating Kiyla results into existing team meetings using the provided ‘Action Planning Template’—a one-page document guiding discussion of domain scores, environmental adjustments, and family partnership goals.

It is important to emphasize that Kiyla’s value increases exponentially when paired with family engagement. Every Kiyla report includes a ‘What This Means for Your Child’ section written at a 5th-grade reading level, with concrete examples: ‘Your child looked at your face when you laughed—that’s an early sign of social connection!’ Rather than labeling, it affirms observable behaviors and invites collaborative next steps. This approach aligns with decades of research showing that caregiver efficacy beliefs predict intervention adherence and child outcomes more strongly than any single assessment score.

As early childhood systems increasingly prioritize measurement-driven improvement, Kiyla offers a rare combination: scientific rigor without logistical burden, precision without exclusion, and structure without rigidity. Its growing adoption—now spanning over 4,100 sites in 42 U.S. states and three Canadian provinces—reflects not marketing momentum but measurable improvements in identification accuracy, referral efficiency, and instructional responsiveness.

The trajectory of Kiyla underscores a broader evolution in early childhood assessment: away from static, clinic-bound evaluations and toward dynamic, relationship-centered measurement. Its emphasis on what children *do*—not what they lack—resonates with contemporary neurodiversity-affirming frameworks. When a toddler stacks four blocks independently, Kiyla documents that achievement with fidelity; when a child uses a gaze shift to request a toy, Kiyla captures that communicative intent without requiring verbal output. In doing so, it honors developmental diversity while maintaining scientific accountability.

For district leaders, Kiyla’s scalability is proven: the Dallas Independent School District implemented it across 112 pre-K campuses in 2023, achieving 98% staff certification compliance within six months using embedded coaching cycles and stipends for lead teachers ($25/hour for 10 hours of mentorship). Their data showed a 29% increase in teacher-reported confidence in recognizing developmental cues—a critical precursor to timely intervention.

Ultimately, Kiyla’s contribution extends beyond metrics. It reshapes professional identity—transforming educators from passive observers into skilled developmental analysts, and clinicians from evaluators into collaborative partners. Its quiet power lies in making developmental science accessible, actionable, and humane—one observed behavior, one calibrated rating, one supported child at a time.

• Kiyla is validated for children aged 12–48 months; no items extend beyond 48 months.
• Standardization sample size: 2,156 children across 14 states.
• Inter-rater reliability: κ = 0.92 (Cohen’s kappa) for primary domains.
• Minimum space requirement: 2.5 m × 2.5 m (8.2 ft × 8.2 ft).
• Digital platform uptime: 99.98% over past 12 months (per AWS status dashboard).

1. Complete Kiyla Foundations Course (4 hours).
2. Submit two video administrations for trainer review (max 72-hour turnaround).
3. Pass live calibration session (observing 3 standardized scenarios).
4. Receive certification badge and access to Kiyla Connect portal.
5. Maintain certification with biennial 90-minute renewal module.

Research continues to affirm Kiyla’s role as a catalyst for system-level change. A 2024 policy brief from the BUILD Initiative documented that states adopting Kiyla as a universal screener saw a 17% average increase in federal Part C funding utilization—indicating better alignment between identification and service provision. This isn’t merely about better tools; it’s about building infrastructure where every child’s developmental story is seen, measured with care, and responded to with intentionality.

As new cohorts of toddlers engage with Kiyla’s playful tasks—rolling balls, naming pictures, stacking blocks—the tool quietly advances a deeper mission: to ensure that developmental surveillance serves not as a filter, but as a bridge—to services, to understanding, and to belonging.